Your Zap stopped running and nobody told you

Why Zaps fail silently

Zapier's error notifications cover runs that start and then fail. Several common failures never start a run, so there's nothing to report:

• Expired or revoked connections. The trigger app can't be polled, so no new data is seen.
• The Zap was turned off — by a teammate, by Zapier after repeated errors, or after a plan or task-limit change.
• The trigger went quiet. The form, inbox or spreadsheet that feeds the Zap stopped receiving data (a broken embed, a changed filter, a renamed sheet).
• Filters and paths quietly stop everything downstream after a field changes.

As one Zapier Community member put it: *"I've been burned twice, both times a customer told me before Zapier did."*

The fix: watch the outcome, not the error

Instead of waiting for an error, expect a result on a schedule and get alerted when it doesn't arrive. That's a heartbeat (also called a dead man's switch):

1. Create a watch in Did It Run? with the schedule the Zap should keep — "every weekday by 9:15am", or "at least once a day" for event-driven Zaps.
2. Add one final step to the Zap: Webhooks by Zapier → POST to your watch's ping URL. Map a summary field (or the number of records processed) into the body so sanity checks have something to look at. Webhooks by Zapier and multi-step Zaps need a paid Zapier plan or Zapier's free trial.
3. If the ping doesn't arrive by the deadline, you get an email (or Slack, Discord or Telegram) alert. When it starts again, you get a recovery notice.

Catch Zaps that run but do nothing

A Zap can "succeed" with empty fields. Send the useful output in the final step and add a sanity check — a minimum size, a minimum item count, or a keyword that must appear. Blank runs raise an empty alert.

One morning roll-call for every Zap

With many Zaps (or many clients), the daily roll-call digest tells you in one email what ran, what's missing, and what came back empty — across Zapier, Make, n8n, ChatGPT tasks and cron.